RESEARCH ARTICLE
A M E R I C A N J O U R N A L O F B OTA N Y
EFFECTS OF ARBUSCULAR MYCORRHIZAL FUNGI AND MATERNAL PLANT SEX ON SEED GERMINATION AND EARLY PLANT ESTABLISHMENT1
SANDRA VARGA2 Department of Biological and Environmental Science, University of Jyväskylä, Jyväskylä, Finland • Premise of the study: Arbuscular mycorrhizal fungi usually enhance overall plant performance, yet their effects on seed germination and early plant establishment, crucial steps in plant cycles, are generally overlooked. In gynodioecious species, sexual dimorphism in these traits has been reported, with females producing seeds that germinate at a faster rate than seeds from hermaphrodites. • Methods: Using the gynodioecious plant Geranium sylvaticum, I investigated in a greenhouse experiment whether the presence of arbuscular mycorrhizal spores affects seed germination and early plant establishment, examining at the same time whether the sex of the mother producing the seeds also influences these parameters and whether sex-specific interactions between these two factors exist. • Key results: The presence of arbuscular mycorrhizal spores in the soil decreased seed germination, did not affect plant survival, but did increase plant growth. Moreover, no significant differences in seed traits were detected between the sexes of the plants producing the seeds. • Conclusions: This study demonstrates that arbuscular mycorrhizal fungi may have contrasting effects for plants during early life stages and that mycorrhizal effects can take place even at the precolonization stage. Key words: Geranium sylvaticum; gynodioecy; seedling; sexual dimorphism; arbuscular mycorrhizal fungal spores.
Most plants grow in symbiosis with fungi from the phylum Glomeromycota in their roots, forming the so-called arbuscular mycorrhizal (AM) symbiosis (Wang and Qiu, 2006). The relationship is considered mutualistic: the heterotrophic fungi are fully dependent on the host plant for carbon acquisition and in exchange, fungi uptake and allocate soil nutrients to the plant (Smith and Read, 2008). The importance of AM fungi for plants is widely recognized and well established: because of increased nutrient uptake and other fungal-derived benefits such as improved water acquisition, mycorrhizal plants grow usually better and have increased reproductive output than nonmycorrhizal plants (Smith and Read, 2008; Koide, 2010). Recently, evidence is accumulating about the existence of sex-specific interactions between AM fungi and sexually dimorphic plants (reviewed in Varga, 2010). Among sexually dimorphic systems, gynodioecy is a relatively common genetic polymorphism found in about 7% of
plants (Richards, 1997). In gynodioecious populations, malesterile (i.e., female) plants coexist with hermaphrodite individuals. This sexual polymorphism is genetically determined (e.g., Koelewijn and Van Damme, 1995). Two major types of gynodioecy have been described depending on whether variation for sex is exclusively determined by nuclear genes or cytoplasmic male-sterility genes are also involved (see e.g., Delph and Kelly, 2014 for a recent review). Whether or not cytoplasmic malesterility genes are involved, it is important to model the level of compensation needed for females to coexist with hermaphrodites (Lewis, 1941). In nuclear-cytoplasmic gynodioecy—the most common type reported across plants—females can be maintained by frequency-depending selection given a fecundity advantage over hermaphrodites and/or the cost of restoration of male fertility associated with carrying nuclear restorer genes (Bailey et al., 2003; Dufaÿ et al., 2008). The female advantage can be achieved through different mechanisms including inbreeding depression avoidance, resource compensation mechanisms, or with less detrimental interactions with pathogens and herbivores (reviewed in Dufaÿ and Billard, 2012). For some gynodioecious species, individuals with intermediate sexual phenotypes have been reported (for a list of cases see Koelewijn and van Damme, 1996). Plants with intermediate sex expression include plants with a mixture of hermaphroditic and female flowers (e.g., gynomonoecious plants) and individuals displaying variability in the number of functional anthers among flowers. The existence of intermediate phenotypes has been proposed to be the result of a polygenic restoration of
1 Manuscript received 15 August 2014; revision accepted 30 January 2015. I thank G. Francini and T. Savolainen for taking care of the plants, M. Vestberg for spore identification, and R. Vega-Frutis and two anonymous reviewers for kindly commenting on a previous draft of the manuscript. This study was financially supported by the Academy of Finland (project number 250911). 2 Corresponding author e-mail: [email protected]; sandravarga30@ hotmail.com
doi:10.3732/ajb.1400361
American Journal of Botany 102(3): 358–366, 2015; http://www.amjbot.org/ © 2015 Botanical Society of America
358
VA R G A — S E E D G E R M I N AT I O N A S A F F E C T E D B Y A M F U N G I A N D M AT E R N A L P L A N T S E X • V O L . 1 0 2 , N O. 3 M A R C H 2 0 1 5
male fertility in hermaphrodites (Koelewijn and van Damme, 1996; Ehlers et al., 2005; Dufaÿ et al., 2008). Fitness comparisons among the three sexes have been performed for very few species (reviewed in Varga and Kytöviita, 2014); and to the best knowledge of the author, no study has investigated seed germination in intermediate plants. Recent findings suggest that the relationship between AM fungi and sexually dimorphic plants may be dependent on the sex of the plant. For instance, distinct AM fungal colonization patterns have been reported in dioecious plants both in greenhouse studies (Eppley et al., 2009) and in the field (Vega-Frutis and Guevara, 2009), which may result in different AM colonization levels between the roots of different sexes (reviewed in Vega-Frutis et al., 2013). These differences in the amount of fungal structures inside roots have been theoretically linked to the reported sex-specific benefits obtained from the root symbioses (Varga and Kytöviita, 2008, 2010). Moreover, sexual dimorphism in seed germination and seedling establishment has been reported for several gynodioecious species (Table 1), with female plants producing seeds that germinate more or faster than seeds from hermaphrodites (Shykoff et al., 2003). However, while the effects of AM fungi on plants are relatively well established in adult plants, few studies have investigated whether AM fungi promote seed germination (Rueda-Puente et al., 2010, Barber et al., 2013, Wu et al., 2014) and to the best knowledge of the author, the effect of AM fungal spores on seed germination has not been examined. At the seedling stage, the presence of AM fungi has been reported to benefit plant
TABLE 1.
• 359
early establishment (e.g., Francis and Read, 1995; van der Heijden, 2004). However, whether the presence of AM propagules (i.e., spores, extraradical hyphae and/or fragments of colonized roots) stimulates seed germination remains largely unexplored. If seed germination in gynodioecious plants is dependent on the sex of the mother (Shykoff et al., 2003), this difference could lead to differences in the rate of AM fungal colonization in the faster germinating seeds and thus it could confer competitive advantage over the slower ones. Therefore, the aim of this study was to investigate whether the presence of AM fungal spores affects seed germination and early plant establishment, examining at the same time whether the sex of the mother producing the seeds also influences these parameters, and whether potential interactions between AM fungi and plant sex exists. To explore plant performance in more detail, plants were harvested when 2, 4, or 6 months old. Specifically, the following hypotheses were tested: (i) the presence of AM fungi will enhance seed germination; (ii) a larger proportion of seeds from female mothers will germinate than seeds produced by intermediates or hermaphrodites; and (iii) mycorrhizal benefits (i.e., plant growth) will become more evident as seedlings get older.
MATERIALS AND METHODS Study species—Geranium sylvaticum L. (Geraniaceae) is a gynodioecious perennial plant with Eurasian distribution found in herb-rich forests, meadows, and along the sides of roads. Populations of G. sylvaticum consist mainly of
Studies in which seed and seedling sexual dimorphism in gynodioecious species were investigated.
Family Asteraceae Boraginaceae Campanulaceae Caryophyllaceae
Colchicaceae Fabaceae Geraniaceae
Hydrophyllaceae Lamiaceae Poaceae Resedaceae Rosaceae Saxifragaceae Thymelaceae
Species Bidens sandvicensis Cirsium arvense Eritrichium aretioides Phacelia linearis Lobelia siphilitica Dianthus sylvestris Schiedea adamantis Schiedea salicaria Silene acaulis
Wurmbea biglandulosa Trifolium hirtum Geranium maculatum Geranium sylvaticum
Phacelia dubia Thymus loscosii Thymus vulgaris Cortaderia richardii Ochradenus baccatus Fragaria virginiana Prunus mahaleb Saxifraga granulata Daphne laureaola
Seed size ns1 ns1,2 ns1,2 F > H1 ns1 ns1 F > H1 F > H1 ns1 ns1 ns1 ns1 ns1 H > F1 ns1 F > H1 ns1 ns1 ns1 ns1 H > F2 H > F1 F > H1 F > H1 F > H1,2 F > H1 ns1 F > H1 F > H1 ns1
Seed nutrient content
Seed germination F>H F>H F>H F>H ns ns ns ns F ≥ HA ns
Seedling growth
Seedling survival F>H
ns3 ns1,4 ns1
F>H ns ns ns ns F>H
ns5,6 F>H F>H F>H ns F>H ns
F > H1 ns1 ns2 F ≥ H1,A
ns F>H ns ns ns ns
ns5 ns ns H ≥ FB F>H F>H F>H F>H F>H F ≥ HB ns ns F>H
ns1 ns1
ns ns
F ≥ HB
ns F>H
Reference Schultz and Ganders, 1996 Lloyd and Myall, 1976 Kay, 1985 Puterbaugh et al., 1997 Eckhart, 1992 Mutikainen and Delph, 1998 Collin et al., 2002 Sakai et al., 1997 Weller and Sakai, 2005 Delph and Mutikainen, 2003 Delph et al., 1999 Keller and Schwaegerle, 2006 Shykoff, 1988 Ramsey and Vaughton, 2002 Molina-Freaner and Jain, 1992 Chang, 2006 Asikainen and Mutikainen, 2003 Varga et al., 2015 Varga et al., 2013 This study Vaarama and Jääskeläinen, 1967 Del Castillo, 1993 Orellana et al., 2005 Gigord et al., 1999 Connor, 1965 Hegazy et al., 2011 Ashman, 1992 Jordano, 1993 Stevens, 1988 Alonso and Herrera, 2001
Notes: F: females; H: hermaphrodites; I: intermediates; ns: nonsignificant differences. 1Mass; 2Size; 3Leaf number; 4Rosette leaf diameter, 5Nitrogen, 6Phosphorus. ADifferences among populations; BDifferences between experiments.
360 • V O L . 1 0 2 , N O. 3 M A R C H 2 0 1 5 • A M E R I C A N J O U R N A L O F B O TA N Y
female and hermaphrodite individuals with 10 functional stamens, but intermediate plants with 1–9 functional stamens per flower or a combination between female, intermediate, and hermaphrodite flowers in different proportions (Varga and Kytöviita, 2014) are frequently reported in most populations (Vaarama and Jääskeläinen, 1967; Asikainen and Mutikainen, 2003; Volkováet al., 2007). The plant has been reported to host AM fungi under field conditions (Korhonen et al., 2004; Varga et al., 2009) even though the identity of the fungal species colonizing the roots is still unknown. Experimental setup—A factorial greenhouse experiment was performed where seeds were germinated from female, intermediate, and hermaphrodite plants with or without AM fungal spores to investigate whether the presence of AM fungal spores and the sex of the mother (where the term “mother” indicates a seed producing plant regardless of sex) that produced the seeds affect seed germination and early plant establishment. Seeds were collected in 2010 from an experimental field established in 2008 near Konnevesi, in central Finland (62°38′N, 26°17′E). The experimental field was an experimental setup designed to evaluate long-term performance in G. sylvaticum. The plants originated from seeds obtained from a controlled handpollination experiment (see Varga et al., 2013 for more details). The experimental population was composed of 448 plants that germinated between January and August 2008 and were planted in September 2008 in a grid of 31 rows of 8–17 individuals each, at interplant distances of 0.6 m. On average, 15 seeds per plant (range 11–21 seeds) were used, collected from 11 female, 11 intermediate, and 11 hermaphrodite individuals randomly selected among the experimental plants, for a total of 501 seeds. After collection, seeds were air dried for one week, weighed, and kept dry at +4°C in paper bags. Before being sown for the experiment, seeds were stratified in February 2012 for six months at +4°C in heat-sterilized sand. The AM fungal spores used in the experiment were extracted from the same experimental field where the seeds were collected by taking 10 soil samples across the field soon after the snow melted in May 2012. Samples were taken using a 6 cm diameter core from the top 20 cm soil horizon. The samples were homogenized and air dried for one week, weighed, and sieved through a 5 mm sieve to remove big soil particles. AM fungal spores were extracted from the soil by wet sieving and decanting using a 0.45 µm sieve. At the same time, a bacterial inoculum was obtained by collecting the water from washing the spores out after filtration through a 5 μm nitrocellulose Millipore filter (Millipore, Molsheim, France) to remove all other AM propagules. A total of 11 different fungal morphotypes were identified under a light microscope, based on morphological characters belonging to the genus Acaulospora (A. capsiculata-like, A. cavernata or scrobiculata-like and Acaulospora sp), Claroideoglomus (C. claroideum and C. etunicatum), Funneliformis (F. caledonium), Glomus (G. hoi and G. sp), Rhizophagus (R. clarus), Scutellospora (S. sp) and Septoglomus (S. constrictum). For the experiment, 10 plastic boxes (27 × 23 × 8 cm) were used for germinating the seeds. The boxes were filled with 5 L of a soil mix consisting of heat-sterilized sand and perlite (7:3, v:v) supplemented with dolomite (5g/L) to raise the pH, and bone meal (1.5 g/L) to serve as a slow-release fertilizer. Half of the boxes were allocated to the AM fungal treatment (referred as “AM boxes” hereafter), while the other half were allocated to the nonmycorrhizal fungal treatment (referred as “NM boxes” hereafter). In AM boxes, 15 spores/g dry weight of soil were added, which reproduced the same natural spore density estimated from the field where they were originally collected, while in NM boxes no spores were added. The bacterial community was restored in all boxes by pipetting 10 mL of the bacterial inoculum. The experiment started in June 2012. In each box, 50 seeds were placed following a 7 × 8 grid to identify germinating seedlings. Seeds were planted at a depth of 1 cm. The position of the seeds was randomized within each box to get a similar number of seeds from each sex and mother genotypes growing into each box and position. The boxes were placed under long-day greenhouse conditions (18 h light) and watered when needed (usually every 2–3 d) until the last harvest took place on December 2013. At the end of the experiment, soil chemical analyses were carried out by Suomen Ympäristöpalvelu Laboratorio—a laboratory certified by FINAS (Finnish Accreditation Service) as T131 (EN ISO/IEC 17025). Methods employed were SFS-EN 13037 (solid to liquid ratio of 1:5 v/v) for soil pH, SFSEN 13039 (loss upon ignition at 550°C for 4 h, for organic matter content), SFS-EN 13654-1 (Kjeldahl method) for total nitrogen, and EPA3051 (microwave-assisted HNO3 extraction) for total potassium and phosphorus. Plant parameters—Seed germination was checked daily and posterior plant survival monitored until individuals were either harvested or if they eventually
died before their appropriate harvest. Three harvests were completed when plants were 2, 4, or 6 months old; individuals were randomly allocated from each experimental treatment as soon as the cotyledons became visible. In each harvest, plants were carefully removed from the box with tweezers—minimizing disturbance to the nearby individuals—washed under running water to remove adhered soil particles, and their mass was divided into above- and belowground. Aboveground dry mass was calculated after drying the sample at 60°C for 48 h. Half of the root system was also dried at 60°C for 48 h to calculate belowground dry weight, and the other half was stored in 50% ethanol and used to assess AM fungal colonization. In all plants, the frequency of AM fungal structures in the roots (hyphae, vesicles/spores, and arbuscules) was determined using the gridline intersect method (McGonigle et al., 1990) at 400× magnification on roots stained with Trypan Blue using the procedure described in Varga et al. (2009). Statistical analyses—All statistical analyses were conducted in R v.3.0.1. (R Development Core Team, 2013). In all cases, a data exploration was first performed to assess the potential best model with which to analyze each type of data (Zuur et al., 2010). Differences in seed mass among mother sexes were analyzed using a general linear model with plant sex as a fixed factor. Data on germination success, and on whether plants survived until the corresponding harvest (yes/no), were analyzed using generalized linear models with a binomial distribution or quasi-binomial distribution when data were over-dispersed. In both cases, mother sex (female, intermediate, hermaphrodite), fungal treatment (AM, NM), and their interaction were included as fixed factors, and box and clone identity as random factors. Differences in variances were determined with the ‘Anova’ function from the ‘Car’ package (Fox and Weisberg, 2011) using type III sum of squares. A general linear mixed model was fitted to the data on plant mass. In the model, the sex of the mother, the fungal treatment, and the harvest (first, second, third) were considered fixed effects, while box and mother clone were included as random factors. Data on plant mass were log-transformed prior to analysis. Similarly, a general linear mixed model was fitted to the data on root colonization by AM structures, but only for plants grown in the presence of AM fungi (as all plants harvested from NM boxes did not show any sign of AM fungal colonization). As before, mother sex, harvest, and their interaction were included as fixed factors, and box and mother clone were included as random factors. Data on the soil chemical parameters were analyzed using independent Welsh’s t tests for each parameter analyzed. Finally, to check whether the probability of seed germination was related to the explanatory variables, Cox proportional hazards mixed models with the ‘coxph’ function from the package ‘survival’ were employed (Therneau, 2013). Mother sex, fungal treatment, and the interaction between these two factors were included as fixed factors and clone as a random factor. After fitting the adequate models, the underlying assumptions of normality and homogeneity of each model employed were checked. Differences among the sexes of the mother were determined using Tukey’s honest significant difference (HSD) with the ‘glht’ function of the ‘multcomp’ package (Hothorn et al., 2008).
RESULTS Seed germination and emergence— Seed mass ranged from 4.1–6.5 mg (5.1 ± 0.1 mg, mean ± SE) and was similar regardless of the mother sex (F2,30 = 0.28, P = 0.76). Seed germination started on 7 July 2012 and lasted until 10 July 2013. During that period, 184 of the 501 sowed seeds germinated. Significant differences in the proportion of germinated seeds were detected between fungal treatments (χ22 = 5.86, P = 0.02; Fig. 1) because more seeds germinated in NM boxes (model estimate = 0.83 ±0.34 SE, z value = 2.42, P = 0.02). The sex of the mother producing the seeds did not affect total seed germination (χ21 = 2.28, P = 0.32) and the interaction between fungal treatment and mother sex was also not statistically significant (χ22 = 0.70, P = 0.70; Fig. 1). Because two clear distinct germinating periods were observed (Fig. 2), the probability of germination was analyzed separately. An Early cohort of seeds germinated between 7 July 2012 and 8 Aug 2012 (66% of the plants), while the rest germinated between 19 Nov 2012 and 30 Aug 2013 (Fig. 2). When
VA R G A — S E E D G E R M I N AT I O N A S A F F E C T E D B Y A M F U N G I A N D M AT E R N A L P L A N T S E X • V O L . 1 0 2 , N O. 3 M A R C H 2 0 1 5
• 361
by females, germination was delayed when sown with AM spores (coef = 1.09, SE = 0.58, z = 1.86, P = 0.04; Fig. 3B).
Fig. 1. Proportion of germinated seeds produced by Female, Intermediate, or Hermaphrodite Geranium sylvaticum plants germinated with AM fungal spores (black bars) or without AM fungi (white bars). Numbers above the bars indicate the number of seeds planted for each sex/fungal treatment categories.
analyzed separately by Early or Late cohort, neither the mother sex nor the fungal treatment affected the probability of germination in the Early cohort (χ22 = 0.73, P = 0.70; χ21 = 0.48, P = 0.49 and χ22 = 3.46, P = 0.18 for the effect of mother sex, fungal treatment, and the interaction between these two factors; Fig. 3A) , but in the Late cohort, there was a significant interaction between fungal treatment and mother sex (χ22 = 2.76, P = 0.25; χ21 = 0.67, P = 0.42 and χ22 = 8.13, P = 0.02 for the effect of mother sex, fungal treatment, and the interaction between these two factors; Fig. 3B). The presence of AM fungi did not affect germination in seeds produced by hermaphrodite (coef = −0.44, SE = 0.42, z = −1.03, P = 0.30) or intermediate mothers (coef = −1.08, SE = 0.58, z = −1.87, P = 0.06), but for seeds produced
Plant survival and mass— From the 184 seeds that germinated, 57% of them survived until their pre-established harvest. Plant survival was the only factor statistically affected by harvest (χ22 = 231.52, P < 0.01; all other P ≥ 0.29): 85% of the seeds survived until their first harvest (2 months) compared to 40% and 41% of seeds that survived until their established harvest at 4 and 6 months harvests, respectively. Total mass in the harvested individuals was significantly affected by the effects of the fungal treatment (F1,95 = 4.02, P = 0.05) and the harvest (F2,95 = 9.06, P < 0.01; Fig. 4), with no significant effect due to the sex of the mother (F2,95 = 0.95, P = 0.39) and a near significant interaction between fungal treatment and the sex of the mother producing the seeds (F2,95 = 2.82, P = 0.07). Plant mass was almost doubled in the presence of mycorrhizal fungi (mean average 169.5 ± 39.3 vs. 89.3 ± 13.5 mg in AM and NM, respectively) and plants achieved the largest mass at the third harvest (6 months old, 194.4 ± 44.0 mg) and the smallest mass when 2 months old (53.3 ± 5.9 mg; Fig. 4). Mycorrhizal colonization— No AM fungal structures were observed in plants grown under NM conditions. Regarding plants grown with AM spores, at each harvest it was observed that most individuals were colonized by AM fungal structures; complete absence of fungal structures was noticed only in two seedlings from the first harvest, and in two more individuals from the second harvest. Considerable amounts of both hyphae and arbuscules were observed in all harvests (Fig. 5), but because of the small amount of vesicles (Fig. 5), data on this fungal structure were not analyzed further. Two month old plants had about 30% of their root system colonized by fungal structures (Fig. 5) and the proportion of root length colonized increased toward the latest harvest for both structures (F2,26 = 9.55, P < 0.01, and F2,31 = 4.80, P = 0.02 for hyphae and arbuscules,
Fig. 2. Histogram showing the frequency of seed germination over the duration of the experiment. The inset graph shows a magnification of the Late cohort.
362 • V O L . 1 0 2 , N O. 3 M A R C H 2 0 1 5 • A M E R I C A N J O U R N A L O F B O TA N Y
Fig. 3. Probability of seed germination based on Kaplan–Meier estimates for (A) Early and (B) Late cohorts of seeds sown with AM fungi (black lines) or without AM fungi (gray lines) from Female (solid lines), Intermediate (dotted lines), and Hermaphrodite (dashed lines) mothers.
respectively; Fig. 5). Mother sex did not statistically affect the amount of root length colonized (F2,29 = 0.41, P = 0.67 and F4,29 = 0.90, P = 0.48 for hyphae and arbuscules, respectively) and no significant interaction between mother sex and harvest were detected (F2,31 = 0.19, P = 0.83 and F4,31 = 0.72, P = 0.59 for hyphae and arbuscules, respectively). Soil chemical parameters— Soil chemical properties measured at the end of the experiment are shown in Table 2. Soil from NM boxes had slightly lower pH and higher organic matter content and total N than soil from AM boxes, whereas no significant statistical differences were detected for total K and P content (Table 2). DISCUSSION In this experiment, seed germination and early plant establishment was examined to see whether they are affected by the presence of mycorrhizal fungal spores and by the sex of the
mother that produced the seeds. The results show that the presence of AM spores in the soil had contrasting effects for the plants depending on the parameter analyzed. The presence of AM fungal spores decreased total seed germination, while plant survival was unaffected and plant mass was increased because of the presence of AM fungi. Contrary to the expectations, the sex of the mother producing the seeds did not affect most of the parameters investigated. AM fungal effects on seed germination and seedling establishment— The positive effects of AM fungi for plant growth (Smith and Read, 2008) and reproduction (Koide, 2010) have been widely reported for decades, at least for juveniles and reproductive adult plants. However, whether AM fungi may also affect seedling establishment, or especially seed germination, is relatively unknown (but see Horton and van der Heijden, 2008 for a review of seedling establishment in the presence of AM hyphal networks), even though these are the most vulnerable stages for determining the establishment of a new individual (Leck et al., 2008 and references there). The current
VA R G A — S E E D G E R M I N AT I O N A S A F F E C T E D B Y A M F U N G I A N D M AT E R N A L P L A N T S E X • V O L . 1 0 2 , N O. 3 M A R C H 2 0 1 5
Fig. 4. Total plant mass (dry weight, in milligrams) in the first (2 month old), second (4 month old) and third (6 month old) harvest in Geranium sylvaticum plants grown in nonmycorrhizal (NM, white bars) or in mycorrhizal soil (AM, black bars). Mean ±SE are indicated (N = 39, 38, and 26 plants for the first, second, and third harvest, respectively).
study showed that the presence of AM fungal spores had a negative effect on seed germination in agreement with previous studies performed with the absence of a pre-established AM hyphal network (Barber et al., 2013, Wu et al., 2014; but see Rueda-Puente et al., 2010). Taken together, these studies suggest that AM fungi may trigger seed germination, but the mechanism remains largely unexplained. In the current study, taking into account that the soil microbial biota was restored in both treatments, there are at least two explanations for this result. First, it would be possible that some pathogen was introduced together with the AM fungal spores even though it is not known (by the author) that any previous studies showing this effect were done. Several bacterial groups occur in association with AM fungal spores. While many of these bacteria are known to enhance seed germination and seedling growth (e.g., HernándezRodríguez et al., 2010), some species could theoretically act
Fig. 5. Proportion of root length colonized by hyphae (black line), arbuscules (gray line), and vesicles (dashed line) in Geranium sylvaticum plants grown in mycorrhizal soil in the first (2 month old), second (4 month old) and third (6 month old) harvest. Mean ±SE are indicated (N = 16, 15, and 11 plants for the first, second, and third harvests, respectively). Data are not shown for plants grown in nonmycorrhizal soil because no fungal structure was observed in any plant.
• 363
as antagonists for seed germination. Even though most of these bacterial groups occur at the spore outer wall layer (and thus they should be represented in both treatments because the bacterial community was restored with the washing water to extract spores from the soil), some bacteria have been found within the spore cytoplasm (Bianciotto et al., 2000). Second, it has been shown that AM spore exudates are able to suppress seed germination, at least from parasitic plants (Louarn et al., 2012). The communication mechanism between plants and AM fungi is complex and many processes are still unknown (see Nadal and Paszkowski, 2013 for a recent review on the topic). To date, germinating spore exudates (GSE), some of which have been recently characterized as lipochito-oligosaccharides (e.g., Maillet et al., 2011), have been shown to stimulate root development and induce gene expression (see e.g., Mukherjee and Ané, 2011). Whether GSE influences seed germination in nonparasitic plants remains to be explored. In natural conditions seedlings become colonized by AM fungi soon after germination, usually within days (e.g., Gay et al., 1982; Allen et al., 1989). Theoretically, this early AM symbiotic formation enables the developing seedlings to acquire water and nutrients from the surrounding soil, giving them an advantage over noncolonized seedlings (Smith and Read, 2008). In this study, seedlings were already colonized by the time the first harvest took place (i.e., 2 months old), and the presence of AM fungi increased plant growth without affecting the probability of survival. Arbuscules (i.e., the place of nutrient exchange between the two partners) were already observed after two months. Engaging in symbiosis with AM fungi is costly for plants in terms of the carbon allocated to the fungus, which may be as high as 20% of the total carbon photosynthates (Jakobsen and Rosendahl, 1990). The difference in mass accumulation between mycorrhizal and nonmycorrhizal individuals became larger as plants grew bigger. Especially in developing seedlings, carbon costs may be too high and may compromise plant growth (e.g., Bethlenfalvay et al., 1982) even though beneficial effects have also been reported (e.g., van der Heijden, 2004). In a similar study, Ronsheim (2012) recently demonstrated age-dependent AM effects by growing Allium individuals for 15 months: after 6 months, plant growth was depressed by AM fungi, not affected at 6 months, but enhanced at 15 months old due to the improved phosphorus acquisition. Whether the results seen in the current study were a consequence of improved phosphorus acquisition, the carbon costs associated with the fungal partner or both could not be distinguished, even though the nutrient contents of the soil used were not expected to be limiting plant growth. Nevertheless, these results also suggest that AM plants could easily outperform nonmycorrhizal plants if mass accumulation would follow the same trend over time. Mother sex effects on seed germination and seedling establishment—One mechanism explaining female maintenance in gynodioecious species is through the production of better quality seeds (i.e., bigger seeds or seeds containing larger amounts of nutrients) compared to hermaphrodites (e.g., Dufaÿ and Billard, 2012). A review of the available literature (Table 1) suggests that this is the case in most species studied to date, because in many species, females produce either bigger seeds or simply seeds that germinate better than those of hermaphrodites. However, and in agreement with previous studies using the same species, differences were not found in seed mass or
364 • V O L . 1 0 2 , N O. 3 M A R C H 2 0 1 5 • A M E R I C A N J O U R N A L O F B O TA N Y
TABLE 2.
NM soil AM soil t value P value
Chemical properties of the soil used in the experiment (mean ± SE are indicated, N = 10) and results from the two-sample Welsh’s t tests. pH
OM (%)
N (mg/Kg)
K (mg/Kg)
P (mg/Kg)
7.46 ± 0.07 7.86 ± 0.02 −5.55
A M E R I C A N J O U R N A L O F B OTA N Y
EFFECTS OF ARBUSCULAR MYCORRHIZAL FUNGI AND MATERNAL PLANT SEX ON SEED GERMINATION AND EARLY PLANT ESTABLISHMENT1
SANDRA VARGA2 Department of Biological and Environmental Science, University of Jyväskylä, Jyväskylä, Finland • Premise of the study: Arbuscular mycorrhizal fungi usually enhance overall plant performance, yet their effects on seed germination and early plant establishment, crucial steps in plant cycles, are generally overlooked. In gynodioecious species, sexual dimorphism in these traits has been reported, with females producing seeds that germinate at a faster rate than seeds from hermaphrodites. • Methods: Using the gynodioecious plant Geranium sylvaticum, I investigated in a greenhouse experiment whether the presence of arbuscular mycorrhizal spores affects seed germination and early plant establishment, examining at the same time whether the sex of the mother producing the seeds also influences these parameters and whether sex-specific interactions between these two factors exist. • Key results: The presence of arbuscular mycorrhizal spores in the soil decreased seed germination, did not affect plant survival, but did increase plant growth. Moreover, no significant differences in seed traits were detected between the sexes of the plants producing the seeds. • Conclusions: This study demonstrates that arbuscular mycorrhizal fungi may have contrasting effects for plants during early life stages and that mycorrhizal effects can take place even at the precolonization stage. Key words: Geranium sylvaticum; gynodioecy; seedling; sexual dimorphism; arbuscular mycorrhizal fungal spores.
Most plants grow in symbiosis with fungi from the phylum Glomeromycota in their roots, forming the so-called arbuscular mycorrhizal (AM) symbiosis (Wang and Qiu, 2006). The relationship is considered mutualistic: the heterotrophic fungi are fully dependent on the host plant for carbon acquisition and in exchange, fungi uptake and allocate soil nutrients to the plant (Smith and Read, 2008). The importance of AM fungi for plants is widely recognized and well established: because of increased nutrient uptake and other fungal-derived benefits such as improved water acquisition, mycorrhizal plants grow usually better and have increased reproductive output than nonmycorrhizal plants (Smith and Read, 2008; Koide, 2010). Recently, evidence is accumulating about the existence of sex-specific interactions between AM fungi and sexually dimorphic plants (reviewed in Varga, 2010). Among sexually dimorphic systems, gynodioecy is a relatively common genetic polymorphism found in about 7% of
plants (Richards, 1997). In gynodioecious populations, malesterile (i.e., female) plants coexist with hermaphrodite individuals. This sexual polymorphism is genetically determined (e.g., Koelewijn and Van Damme, 1995). Two major types of gynodioecy have been described depending on whether variation for sex is exclusively determined by nuclear genes or cytoplasmic male-sterility genes are also involved (see e.g., Delph and Kelly, 2014 for a recent review). Whether or not cytoplasmic malesterility genes are involved, it is important to model the level of compensation needed for females to coexist with hermaphrodites (Lewis, 1941). In nuclear-cytoplasmic gynodioecy—the most common type reported across plants—females can be maintained by frequency-depending selection given a fecundity advantage over hermaphrodites and/or the cost of restoration of male fertility associated with carrying nuclear restorer genes (Bailey et al., 2003; Dufaÿ et al., 2008). The female advantage can be achieved through different mechanisms including inbreeding depression avoidance, resource compensation mechanisms, or with less detrimental interactions with pathogens and herbivores (reviewed in Dufaÿ and Billard, 2012). For some gynodioecious species, individuals with intermediate sexual phenotypes have been reported (for a list of cases see Koelewijn and van Damme, 1996). Plants with intermediate sex expression include plants with a mixture of hermaphroditic and female flowers (e.g., gynomonoecious plants) and individuals displaying variability in the number of functional anthers among flowers. The existence of intermediate phenotypes has been proposed to be the result of a polygenic restoration of
1 Manuscript received 15 August 2014; revision accepted 30 January 2015. I thank G. Francini and T. Savolainen for taking care of the plants, M. Vestberg for spore identification, and R. Vega-Frutis and two anonymous reviewers for kindly commenting on a previous draft of the manuscript. This study was financially supported by the Academy of Finland (project number 250911). 2 Corresponding author e-mail: [email protected]; sandravarga30@ hotmail.com
doi:10.3732/ajb.1400361
American Journal of Botany 102(3): 358–366, 2015; http://www.amjbot.org/ © 2015 Botanical Society of America
358
VA R G A — S E E D G E R M I N AT I O N A S A F F E C T E D B Y A M F U N G I A N D M AT E R N A L P L A N T S E X • V O L . 1 0 2 , N O. 3 M A R C H 2 0 1 5
male fertility in hermaphrodites (Koelewijn and van Damme, 1996; Ehlers et al., 2005; Dufaÿ et al., 2008). Fitness comparisons among the three sexes have been performed for very few species (reviewed in Varga and Kytöviita, 2014); and to the best knowledge of the author, no study has investigated seed germination in intermediate plants. Recent findings suggest that the relationship between AM fungi and sexually dimorphic plants may be dependent on the sex of the plant. For instance, distinct AM fungal colonization patterns have been reported in dioecious plants both in greenhouse studies (Eppley et al., 2009) and in the field (Vega-Frutis and Guevara, 2009), which may result in different AM colonization levels between the roots of different sexes (reviewed in Vega-Frutis et al., 2013). These differences in the amount of fungal structures inside roots have been theoretically linked to the reported sex-specific benefits obtained from the root symbioses (Varga and Kytöviita, 2008, 2010). Moreover, sexual dimorphism in seed germination and seedling establishment has been reported for several gynodioecious species (Table 1), with female plants producing seeds that germinate more or faster than seeds from hermaphrodites (Shykoff et al., 2003). However, while the effects of AM fungi on plants are relatively well established in adult plants, few studies have investigated whether AM fungi promote seed germination (Rueda-Puente et al., 2010, Barber et al., 2013, Wu et al., 2014) and to the best knowledge of the author, the effect of AM fungal spores on seed germination has not been examined. At the seedling stage, the presence of AM fungi has been reported to benefit plant
TABLE 1.
• 359
early establishment (e.g., Francis and Read, 1995; van der Heijden, 2004). However, whether the presence of AM propagules (i.e., spores, extraradical hyphae and/or fragments of colonized roots) stimulates seed germination remains largely unexplored. If seed germination in gynodioecious plants is dependent on the sex of the mother (Shykoff et al., 2003), this difference could lead to differences in the rate of AM fungal colonization in the faster germinating seeds and thus it could confer competitive advantage over the slower ones. Therefore, the aim of this study was to investigate whether the presence of AM fungal spores affects seed germination and early plant establishment, examining at the same time whether the sex of the mother producing the seeds also influences these parameters, and whether potential interactions between AM fungi and plant sex exists. To explore plant performance in more detail, plants were harvested when 2, 4, or 6 months old. Specifically, the following hypotheses were tested: (i) the presence of AM fungi will enhance seed germination; (ii) a larger proportion of seeds from female mothers will germinate than seeds produced by intermediates or hermaphrodites; and (iii) mycorrhizal benefits (i.e., plant growth) will become more evident as seedlings get older.
MATERIALS AND METHODS Study species—Geranium sylvaticum L. (Geraniaceae) is a gynodioecious perennial plant with Eurasian distribution found in herb-rich forests, meadows, and along the sides of roads. Populations of G. sylvaticum consist mainly of
Studies in which seed and seedling sexual dimorphism in gynodioecious species were investigated.
Family Asteraceae Boraginaceae Campanulaceae Caryophyllaceae
Colchicaceae Fabaceae Geraniaceae
Hydrophyllaceae Lamiaceae Poaceae Resedaceae Rosaceae Saxifragaceae Thymelaceae
Species Bidens sandvicensis Cirsium arvense Eritrichium aretioides Phacelia linearis Lobelia siphilitica Dianthus sylvestris Schiedea adamantis Schiedea salicaria Silene acaulis
Wurmbea biglandulosa Trifolium hirtum Geranium maculatum Geranium sylvaticum
Phacelia dubia Thymus loscosii Thymus vulgaris Cortaderia richardii Ochradenus baccatus Fragaria virginiana Prunus mahaleb Saxifraga granulata Daphne laureaola
Seed size ns1 ns1,2 ns1,2 F > H1 ns1 ns1 F > H1 F > H1 ns1 ns1 ns1 ns1 ns1 H > F1 ns1 F > H1 ns1 ns1 ns1 ns1 H > F2 H > F1 F > H1 F > H1 F > H1,2 F > H1 ns1 F > H1 F > H1 ns1
Seed nutrient content
Seed germination F>H F>H F>H F>H ns ns ns ns F ≥ HA ns
Seedling growth
Seedling survival F>H
ns3 ns1,4 ns1
F>H ns ns ns ns F>H
ns5,6 F>H F>H F>H ns F>H ns
F > H1 ns1 ns2 F ≥ H1,A
ns F>H ns ns ns ns
ns5 ns ns H ≥ FB F>H F>H F>H F>H F>H F ≥ HB ns ns F>H
ns1 ns1
ns ns
F ≥ HB
ns F>H
Reference Schultz and Ganders, 1996 Lloyd and Myall, 1976 Kay, 1985 Puterbaugh et al., 1997 Eckhart, 1992 Mutikainen and Delph, 1998 Collin et al., 2002 Sakai et al., 1997 Weller and Sakai, 2005 Delph and Mutikainen, 2003 Delph et al., 1999 Keller and Schwaegerle, 2006 Shykoff, 1988 Ramsey and Vaughton, 2002 Molina-Freaner and Jain, 1992 Chang, 2006 Asikainen and Mutikainen, 2003 Varga et al., 2015 Varga et al., 2013 This study Vaarama and Jääskeläinen, 1967 Del Castillo, 1993 Orellana et al., 2005 Gigord et al., 1999 Connor, 1965 Hegazy et al., 2011 Ashman, 1992 Jordano, 1993 Stevens, 1988 Alonso and Herrera, 2001
Notes: F: females; H: hermaphrodites; I: intermediates; ns: nonsignificant differences. 1Mass; 2Size; 3Leaf number; 4Rosette leaf diameter, 5Nitrogen, 6Phosphorus. ADifferences among populations; BDifferences between experiments.
360 • V O L . 1 0 2 , N O. 3 M A R C H 2 0 1 5 • A M E R I C A N J O U R N A L O F B O TA N Y
female and hermaphrodite individuals with 10 functional stamens, but intermediate plants with 1–9 functional stamens per flower or a combination between female, intermediate, and hermaphrodite flowers in different proportions (Varga and Kytöviita, 2014) are frequently reported in most populations (Vaarama and Jääskeläinen, 1967; Asikainen and Mutikainen, 2003; Volkováet al., 2007). The plant has been reported to host AM fungi under field conditions (Korhonen et al., 2004; Varga et al., 2009) even though the identity of the fungal species colonizing the roots is still unknown. Experimental setup—A factorial greenhouse experiment was performed where seeds were germinated from female, intermediate, and hermaphrodite plants with or without AM fungal spores to investigate whether the presence of AM fungal spores and the sex of the mother (where the term “mother” indicates a seed producing plant regardless of sex) that produced the seeds affect seed germination and early plant establishment. Seeds were collected in 2010 from an experimental field established in 2008 near Konnevesi, in central Finland (62°38′N, 26°17′E). The experimental field was an experimental setup designed to evaluate long-term performance in G. sylvaticum. The plants originated from seeds obtained from a controlled handpollination experiment (see Varga et al., 2013 for more details). The experimental population was composed of 448 plants that germinated between January and August 2008 and were planted in September 2008 in a grid of 31 rows of 8–17 individuals each, at interplant distances of 0.6 m. On average, 15 seeds per plant (range 11–21 seeds) were used, collected from 11 female, 11 intermediate, and 11 hermaphrodite individuals randomly selected among the experimental plants, for a total of 501 seeds. After collection, seeds were air dried for one week, weighed, and kept dry at +4°C in paper bags. Before being sown for the experiment, seeds were stratified in February 2012 for six months at +4°C in heat-sterilized sand. The AM fungal spores used in the experiment were extracted from the same experimental field where the seeds were collected by taking 10 soil samples across the field soon after the snow melted in May 2012. Samples were taken using a 6 cm diameter core from the top 20 cm soil horizon. The samples were homogenized and air dried for one week, weighed, and sieved through a 5 mm sieve to remove big soil particles. AM fungal spores were extracted from the soil by wet sieving and decanting using a 0.45 µm sieve. At the same time, a bacterial inoculum was obtained by collecting the water from washing the spores out after filtration through a 5 μm nitrocellulose Millipore filter (Millipore, Molsheim, France) to remove all other AM propagules. A total of 11 different fungal morphotypes were identified under a light microscope, based on morphological characters belonging to the genus Acaulospora (A. capsiculata-like, A. cavernata or scrobiculata-like and Acaulospora sp), Claroideoglomus (C. claroideum and C. etunicatum), Funneliformis (F. caledonium), Glomus (G. hoi and G. sp), Rhizophagus (R. clarus), Scutellospora (S. sp) and Septoglomus (S. constrictum). For the experiment, 10 plastic boxes (27 × 23 × 8 cm) were used for germinating the seeds. The boxes were filled with 5 L of a soil mix consisting of heat-sterilized sand and perlite (7:3, v:v) supplemented with dolomite (5g/L) to raise the pH, and bone meal (1.5 g/L) to serve as a slow-release fertilizer. Half of the boxes were allocated to the AM fungal treatment (referred as “AM boxes” hereafter), while the other half were allocated to the nonmycorrhizal fungal treatment (referred as “NM boxes” hereafter). In AM boxes, 15 spores/g dry weight of soil were added, which reproduced the same natural spore density estimated from the field where they were originally collected, while in NM boxes no spores were added. The bacterial community was restored in all boxes by pipetting 10 mL of the bacterial inoculum. The experiment started in June 2012. In each box, 50 seeds were placed following a 7 × 8 grid to identify germinating seedlings. Seeds were planted at a depth of 1 cm. The position of the seeds was randomized within each box to get a similar number of seeds from each sex and mother genotypes growing into each box and position. The boxes were placed under long-day greenhouse conditions (18 h light) and watered when needed (usually every 2–3 d) until the last harvest took place on December 2013. At the end of the experiment, soil chemical analyses were carried out by Suomen Ympäristöpalvelu Laboratorio—a laboratory certified by FINAS (Finnish Accreditation Service) as T131 (EN ISO/IEC 17025). Methods employed were SFS-EN 13037 (solid to liquid ratio of 1:5 v/v) for soil pH, SFSEN 13039 (loss upon ignition at 550°C for 4 h, for organic matter content), SFS-EN 13654-1 (Kjeldahl method) for total nitrogen, and EPA3051 (microwave-assisted HNO3 extraction) for total potassium and phosphorus. Plant parameters—Seed germination was checked daily and posterior plant survival monitored until individuals were either harvested or if they eventually
died before their appropriate harvest. Three harvests were completed when plants were 2, 4, or 6 months old; individuals were randomly allocated from each experimental treatment as soon as the cotyledons became visible. In each harvest, plants were carefully removed from the box with tweezers—minimizing disturbance to the nearby individuals—washed under running water to remove adhered soil particles, and their mass was divided into above- and belowground. Aboveground dry mass was calculated after drying the sample at 60°C for 48 h. Half of the root system was also dried at 60°C for 48 h to calculate belowground dry weight, and the other half was stored in 50% ethanol and used to assess AM fungal colonization. In all plants, the frequency of AM fungal structures in the roots (hyphae, vesicles/spores, and arbuscules) was determined using the gridline intersect method (McGonigle et al., 1990) at 400× magnification on roots stained with Trypan Blue using the procedure described in Varga et al. (2009). Statistical analyses—All statistical analyses were conducted in R v.3.0.1. (R Development Core Team, 2013). In all cases, a data exploration was first performed to assess the potential best model with which to analyze each type of data (Zuur et al., 2010). Differences in seed mass among mother sexes were analyzed using a general linear model with plant sex as a fixed factor. Data on germination success, and on whether plants survived until the corresponding harvest (yes/no), were analyzed using generalized linear models with a binomial distribution or quasi-binomial distribution when data were over-dispersed. In both cases, mother sex (female, intermediate, hermaphrodite), fungal treatment (AM, NM), and their interaction were included as fixed factors, and box and clone identity as random factors. Differences in variances were determined with the ‘Anova’ function from the ‘Car’ package (Fox and Weisberg, 2011) using type III sum of squares. A general linear mixed model was fitted to the data on plant mass. In the model, the sex of the mother, the fungal treatment, and the harvest (first, second, third) were considered fixed effects, while box and mother clone were included as random factors. Data on plant mass were log-transformed prior to analysis. Similarly, a general linear mixed model was fitted to the data on root colonization by AM structures, but only for plants grown in the presence of AM fungi (as all plants harvested from NM boxes did not show any sign of AM fungal colonization). As before, mother sex, harvest, and their interaction were included as fixed factors, and box and mother clone were included as random factors. Data on the soil chemical parameters were analyzed using independent Welsh’s t tests for each parameter analyzed. Finally, to check whether the probability of seed germination was related to the explanatory variables, Cox proportional hazards mixed models with the ‘coxph’ function from the package ‘survival’ were employed (Therneau, 2013). Mother sex, fungal treatment, and the interaction between these two factors were included as fixed factors and clone as a random factor. After fitting the adequate models, the underlying assumptions of normality and homogeneity of each model employed were checked. Differences among the sexes of the mother were determined using Tukey’s honest significant difference (HSD) with the ‘glht’ function of the ‘multcomp’ package (Hothorn et al., 2008).
RESULTS Seed germination and emergence— Seed mass ranged from 4.1–6.5 mg (5.1 ± 0.1 mg, mean ± SE) and was similar regardless of the mother sex (F2,30 = 0.28, P = 0.76). Seed germination started on 7 July 2012 and lasted until 10 July 2013. During that period, 184 of the 501 sowed seeds germinated. Significant differences in the proportion of germinated seeds were detected between fungal treatments (χ22 = 5.86, P = 0.02; Fig. 1) because more seeds germinated in NM boxes (model estimate = 0.83 ±0.34 SE, z value = 2.42, P = 0.02). The sex of the mother producing the seeds did not affect total seed germination (χ21 = 2.28, P = 0.32) and the interaction between fungal treatment and mother sex was also not statistically significant (χ22 = 0.70, P = 0.70; Fig. 1). Because two clear distinct germinating periods were observed (Fig. 2), the probability of germination was analyzed separately. An Early cohort of seeds germinated between 7 July 2012 and 8 Aug 2012 (66% of the plants), while the rest germinated between 19 Nov 2012 and 30 Aug 2013 (Fig. 2). When
VA R G A — S E E D G E R M I N AT I O N A S A F F E C T E D B Y A M F U N G I A N D M AT E R N A L P L A N T S E X • V O L . 1 0 2 , N O. 3 M A R C H 2 0 1 5
• 361
by females, germination was delayed when sown with AM spores (coef = 1.09, SE = 0.58, z = 1.86, P = 0.04; Fig. 3B).
Fig. 1. Proportion of germinated seeds produced by Female, Intermediate, or Hermaphrodite Geranium sylvaticum plants germinated with AM fungal spores (black bars) or without AM fungi (white bars). Numbers above the bars indicate the number of seeds planted for each sex/fungal treatment categories.
analyzed separately by Early or Late cohort, neither the mother sex nor the fungal treatment affected the probability of germination in the Early cohort (χ22 = 0.73, P = 0.70; χ21 = 0.48, P = 0.49 and χ22 = 3.46, P = 0.18 for the effect of mother sex, fungal treatment, and the interaction between these two factors; Fig. 3A) , but in the Late cohort, there was a significant interaction between fungal treatment and mother sex (χ22 = 2.76, P = 0.25; χ21 = 0.67, P = 0.42 and χ22 = 8.13, P = 0.02 for the effect of mother sex, fungal treatment, and the interaction between these two factors; Fig. 3B). The presence of AM fungi did not affect germination in seeds produced by hermaphrodite (coef = −0.44, SE = 0.42, z = −1.03, P = 0.30) or intermediate mothers (coef = −1.08, SE = 0.58, z = −1.87, P = 0.06), but for seeds produced
Plant survival and mass— From the 184 seeds that germinated, 57% of them survived until their pre-established harvest. Plant survival was the only factor statistically affected by harvest (χ22 = 231.52, P < 0.01; all other P ≥ 0.29): 85% of the seeds survived until their first harvest (2 months) compared to 40% and 41% of seeds that survived until their established harvest at 4 and 6 months harvests, respectively. Total mass in the harvested individuals was significantly affected by the effects of the fungal treatment (F1,95 = 4.02, P = 0.05) and the harvest (F2,95 = 9.06, P < 0.01; Fig. 4), with no significant effect due to the sex of the mother (F2,95 = 0.95, P = 0.39) and a near significant interaction between fungal treatment and the sex of the mother producing the seeds (F2,95 = 2.82, P = 0.07). Plant mass was almost doubled in the presence of mycorrhizal fungi (mean average 169.5 ± 39.3 vs. 89.3 ± 13.5 mg in AM and NM, respectively) and plants achieved the largest mass at the third harvest (6 months old, 194.4 ± 44.0 mg) and the smallest mass when 2 months old (53.3 ± 5.9 mg; Fig. 4). Mycorrhizal colonization— No AM fungal structures were observed in plants grown under NM conditions. Regarding plants grown with AM spores, at each harvest it was observed that most individuals were colonized by AM fungal structures; complete absence of fungal structures was noticed only in two seedlings from the first harvest, and in two more individuals from the second harvest. Considerable amounts of both hyphae and arbuscules were observed in all harvests (Fig. 5), but because of the small amount of vesicles (Fig. 5), data on this fungal structure were not analyzed further. Two month old plants had about 30% of their root system colonized by fungal structures (Fig. 5) and the proportion of root length colonized increased toward the latest harvest for both structures (F2,26 = 9.55, P < 0.01, and F2,31 = 4.80, P = 0.02 for hyphae and arbuscules,
Fig. 2. Histogram showing the frequency of seed germination over the duration of the experiment. The inset graph shows a magnification of the Late cohort.
362 • V O L . 1 0 2 , N O. 3 M A R C H 2 0 1 5 • A M E R I C A N J O U R N A L O F B O TA N Y
Fig. 3. Probability of seed germination based on Kaplan–Meier estimates for (A) Early and (B) Late cohorts of seeds sown with AM fungi (black lines) or without AM fungi (gray lines) from Female (solid lines), Intermediate (dotted lines), and Hermaphrodite (dashed lines) mothers.
respectively; Fig. 5). Mother sex did not statistically affect the amount of root length colonized (F2,29 = 0.41, P = 0.67 and F4,29 = 0.90, P = 0.48 for hyphae and arbuscules, respectively) and no significant interaction between mother sex and harvest were detected (F2,31 = 0.19, P = 0.83 and F4,31 = 0.72, P = 0.59 for hyphae and arbuscules, respectively). Soil chemical parameters— Soil chemical properties measured at the end of the experiment are shown in Table 2. Soil from NM boxes had slightly lower pH and higher organic matter content and total N than soil from AM boxes, whereas no significant statistical differences were detected for total K and P content (Table 2). DISCUSSION In this experiment, seed germination and early plant establishment was examined to see whether they are affected by the presence of mycorrhizal fungal spores and by the sex of the
mother that produced the seeds. The results show that the presence of AM spores in the soil had contrasting effects for the plants depending on the parameter analyzed. The presence of AM fungal spores decreased total seed germination, while plant survival was unaffected and plant mass was increased because of the presence of AM fungi. Contrary to the expectations, the sex of the mother producing the seeds did not affect most of the parameters investigated. AM fungal effects on seed germination and seedling establishment— The positive effects of AM fungi for plant growth (Smith and Read, 2008) and reproduction (Koide, 2010) have been widely reported for decades, at least for juveniles and reproductive adult plants. However, whether AM fungi may also affect seedling establishment, or especially seed germination, is relatively unknown (but see Horton and van der Heijden, 2008 for a review of seedling establishment in the presence of AM hyphal networks), even though these are the most vulnerable stages for determining the establishment of a new individual (Leck et al., 2008 and references there). The current
VA R G A — S E E D G E R M I N AT I O N A S A F F E C T E D B Y A M F U N G I A N D M AT E R N A L P L A N T S E X • V O L . 1 0 2 , N O. 3 M A R C H 2 0 1 5
Fig. 4. Total plant mass (dry weight, in milligrams) in the first (2 month old), second (4 month old) and third (6 month old) harvest in Geranium sylvaticum plants grown in nonmycorrhizal (NM, white bars) or in mycorrhizal soil (AM, black bars). Mean ±SE are indicated (N = 39, 38, and 26 plants for the first, second, and third harvest, respectively).
study showed that the presence of AM fungal spores had a negative effect on seed germination in agreement with previous studies performed with the absence of a pre-established AM hyphal network (Barber et al., 2013, Wu et al., 2014; but see Rueda-Puente et al., 2010). Taken together, these studies suggest that AM fungi may trigger seed germination, but the mechanism remains largely unexplained. In the current study, taking into account that the soil microbial biota was restored in both treatments, there are at least two explanations for this result. First, it would be possible that some pathogen was introduced together with the AM fungal spores even though it is not known (by the author) that any previous studies showing this effect were done. Several bacterial groups occur in association with AM fungal spores. While many of these bacteria are known to enhance seed germination and seedling growth (e.g., HernándezRodríguez et al., 2010), some species could theoretically act
Fig. 5. Proportion of root length colonized by hyphae (black line), arbuscules (gray line), and vesicles (dashed line) in Geranium sylvaticum plants grown in mycorrhizal soil in the first (2 month old), second (4 month old) and third (6 month old) harvest. Mean ±SE are indicated (N = 16, 15, and 11 plants for the first, second, and third harvests, respectively). Data are not shown for plants grown in nonmycorrhizal soil because no fungal structure was observed in any plant.
• 363
as antagonists for seed germination. Even though most of these bacterial groups occur at the spore outer wall layer (and thus they should be represented in both treatments because the bacterial community was restored with the washing water to extract spores from the soil), some bacteria have been found within the spore cytoplasm (Bianciotto et al., 2000). Second, it has been shown that AM spore exudates are able to suppress seed germination, at least from parasitic plants (Louarn et al., 2012). The communication mechanism between plants and AM fungi is complex and many processes are still unknown (see Nadal and Paszkowski, 2013 for a recent review on the topic). To date, germinating spore exudates (GSE), some of which have been recently characterized as lipochito-oligosaccharides (e.g., Maillet et al., 2011), have been shown to stimulate root development and induce gene expression (see e.g., Mukherjee and Ané, 2011). Whether GSE influences seed germination in nonparasitic plants remains to be explored. In natural conditions seedlings become colonized by AM fungi soon after germination, usually within days (e.g., Gay et al., 1982; Allen et al., 1989). Theoretically, this early AM symbiotic formation enables the developing seedlings to acquire water and nutrients from the surrounding soil, giving them an advantage over noncolonized seedlings (Smith and Read, 2008). In this study, seedlings were already colonized by the time the first harvest took place (i.e., 2 months old), and the presence of AM fungi increased plant growth without affecting the probability of survival. Arbuscules (i.e., the place of nutrient exchange between the two partners) were already observed after two months. Engaging in symbiosis with AM fungi is costly for plants in terms of the carbon allocated to the fungus, which may be as high as 20% of the total carbon photosynthates (Jakobsen and Rosendahl, 1990). The difference in mass accumulation between mycorrhizal and nonmycorrhizal individuals became larger as plants grew bigger. Especially in developing seedlings, carbon costs may be too high and may compromise plant growth (e.g., Bethlenfalvay et al., 1982) even though beneficial effects have also been reported (e.g., van der Heijden, 2004). In a similar study, Ronsheim (2012) recently demonstrated age-dependent AM effects by growing Allium individuals for 15 months: after 6 months, plant growth was depressed by AM fungi, not affected at 6 months, but enhanced at 15 months old due to the improved phosphorus acquisition. Whether the results seen in the current study were a consequence of improved phosphorus acquisition, the carbon costs associated with the fungal partner or both could not be distinguished, even though the nutrient contents of the soil used were not expected to be limiting plant growth. Nevertheless, these results also suggest that AM plants could easily outperform nonmycorrhizal plants if mass accumulation would follow the same trend over time. Mother sex effects on seed germination and seedling establishment—One mechanism explaining female maintenance in gynodioecious species is through the production of better quality seeds (i.e., bigger seeds or seeds containing larger amounts of nutrients) compared to hermaphrodites (e.g., Dufaÿ and Billard, 2012). A review of the available literature (Table 1) suggests that this is the case in most species studied to date, because in many species, females produce either bigger seeds or simply seeds that germinate better than those of hermaphrodites. However, and in agreement with previous studies using the same species, differences were not found in seed mass or
364 • V O L . 1 0 2 , N O. 3 M A R C H 2 0 1 5 • A M E R I C A N J O U R N A L O F B O TA N Y
TABLE 2.
NM soil AM soil t value P value
Chemical properties of the soil used in the experiment (mean ± SE are indicated, N = 10) and results from the two-sample Welsh’s t tests. pH
OM (%)
N (mg/Kg)
K (mg/Kg)
P (mg/Kg)
7.46 ± 0.07 7.86 ± 0.02 −5.55
